How can I store the bits of a uint64_t inside of a double?

Question

Basically I have a uint64_t whose actual value I do not care about. I need to store it in a double so that I can easily store the bits in an object in R (if you don't know what that is, that's fine, it doesn't matter for the question).

So what I would like is a means of storing the 64 bits of the uint64_t inside of a double and then also convert from the double holding the bits back to the original uint64_t .

I've been banging my head against the wall on this for a quite a bit (is that a pun?) so any and all help is greatly appreciated!!!

Answer 1

As stated in the comments, you can use memcpy to copy the bits from your uint64_t to a double .

This works because the size of a double is the same as the one of an uint64_t (8 bytes).

If you try to display the values, you won't get the same result, though. Indeed a double stores both positive and negative values, with a floating point, whereas an uint64_t is... unsigned and whole.

The binary representation is the same, the interpreted value is different.

Answer 2

in C casting is your friend:

• dbl = * (double*)(&ui64);
• ui64 = * (uint64_t*)(&dbl);

example:

#include <stdint.h>
#include <inttypes.h>
#include <stdio.h>

double dbl_1 = 42.42;
uint64_t ui64_1;

double dbl_2;
uint64_t ui64_2 = 0x404535C28F5C28F6;

void main(void)
{
   ui64_1 = *(uint64_t*)(&dbl_1);
   printf("%" PRIx64 "\n", ui64_1);

   dbl_2 = *(double*)(&ui64_2);
   printf("%f", dbl_2); 
}

Great resource for conversions: IEEE-754 Floating-Point Conversion

Edit: This is undefined behavior per the C standards and although it may work, it is not best practice.